<p>
[quote]
afan, you also assume that there is some sort of correlation between SAT I scores and grades in college coursework- do you have any proof of that claim?
[/quote]
</p>
<p>Wow, Drab, I assumed everyone on CC was familiar with this. There is a huge literature on the subject, going back decades. A review of the literature would go to book length. For starters, you might look at "Black White Test Score Gap", by Jencks and Phillips and "Equity and Excellence in American Higher Education" by Bowen, et al. Both are books that contain original research as well as extensive reviews of the literature.</p>
<p>Once you read these, and a few of the largest studies they mention, which include hundresds of thousands of students at colleges across the country, you will find that no informed person questions whether SAT predicts college grades. The concern is that it does this largely or entirely due to the association between SAT and socioeconomic status, and people from higher SAS groups get better grades. Therefore, critics say that the SAT is unfair because it rewards students for being from high SAS families rather than for exceeding what would be expected from their backgrounds. However, even these critics do not, and cannot, deny that higher SAT scores predict higher grades. The only questions are why, and whether the effect is worth the SAS preference it builds into admissions.</p>
<p>bern, Yes, a more detailed look at grades vs input would take into account college majors. Hard sciences and engineering tend to have lower GPA's, I can cite the data if you want, but this is well established. Here UCB and Stanford are good comparisions, because both have a high proportion of physical science and engineering majors. Stanford actually grants ~37% of degrees in STEM fields, vs ~32% for Berkeley. So distribution into low-gpa majors cannot explain the gap.</p>